Looking back at the recent Society of Vacuum Coaters Annual Technical Conference I think that the most interesting set of papers that I listened to were all about atomic layer deposition. This technology has been around for many years but has generally been used on rigid substrates only. The technique can be used to deposit many coatings and a wide variety of thicknesses. Recently this technique has been identified as being potentially useful for other markets providing it can be converted from depositing only on rigid substrates to being capable of depositing coatings onto rolls of material in a roll-to-roll coating process. There look to be two main routes to this goal, one is to use multiple deposition heads each separated by a buffer zone, the other looks as if it will be a cheaper option and that is to use multiple zones within the vacuum system and sequentially wind the web through the zones multiple times. To deposit a 50 layer thick coating would require 50 deposition heads and so could become a capital expensive system. The second process only needs a winding system with a serpentine path. This would require many rollers to provide the necessary serpentine path but would still be a lower cost option. Plus as the gases would not be pumped out so frequently the material utilisation looks as if it too would be the cheaper option.
There are at least two demonstration systems currently being used for process development and others are already designed and are being built and so this technology looks as if it will become available and delivering product within the next 2 – 5 years.
One primary target for material produced by ALD is the barrier coating market. The results would suggest that coating the same substrate by conventional evaporation or chemical vapour deposition techniques does not produce as good a barrier as can be produced by ALD. The limitation of conventional deposition techniques is the pinholes that are produced because of the debris present on the surface during deposition that is coated but then the debris gets moved and so a pinhole in the coating is left behind. Where ALD wins would appear to be how the coating is able to be more conformal to the surface than the conventional techniques. The monolayer of gas is able to coat the substrate surface right up to the root of where the debris touches the substrate and around the whole of the debris. Conventional techniques tend to have the problem of the debris shadowing the substrate and because of the line-of-sight deposition flux there is a larger uncoated area of the substrate under the debris. The higher deposition pressure of the ALD process allows gas scattering to enable the deposition gas to reach right under the debris and so the uncoated area is only the direct contact area between the debris and substrate. So even for the same amount of debris contamination of the substrate and subsequent movement of a similar number of particles leaving behind the same number of pinholes the exposed area of substrate is expected to be much lower for the ALD process than for the conventional deposition processes.
Coupled to this, it is almost as easy for the ALD process to coat both sides of the film at the same time and so if required the cost of doing double side coating would be almost the same as it could be done at the same speed and at the same time. Only the increased consumption rate of the precursor gases would add to the deposition costs.
All this sounds very positive but there will be significant process expertise required to produce good quality coatings. The cleanliness of the system and management of the winding system will not be trivial. With tens of rollers needed to wind between zones to build up the coating thickness it is important that the rollers remain free running and clean. Any problem with even a single roller can damage the coating, producing scratches or pinholes of both, reducing the barrier performance significantly. This might be both bad news and also good news. If some debris is moved after a few mono-layers of deposition the exposed pinhole would then start to become coated with material. Thus it is likely that many pinholes would be partially coated and only the pinholes produced after the end of the deposition process would be true pinholes all the way through the coating to the substrate. Along with the more conformal nature of the coating this could be part of the reason why the coatings are looking to show a better barrier performance for the same materials and at the same thickness as those produced by conventional methods.
Until there is a system that is large enough to produce production amounts of material that can be sold and converted into packages the true performance of the coated material is just speculation. However the technology and pace of development does make the technology one of the more exciting prospects for producing higher quality barrier coatings as well as being able to reduce the costs of the more difficult ultra-barrier coated film products.